## [ForestFolio: Mapping below the forest canopy](/spotlight/student-project/forestfolio-mapping-below-forest-canopy) ![COMP design team](/sites/default/files/styles/max_480w/public/spotlight-images/2024-04/COMP%20D%2BID.jpg.webp?itok=7zw-iamy) ## Kamran Alam, Vicky Chen, Manvir Dhami, Zoeb Gaurani, Harman Sihota and Yitong Tang Share- [ ](https://www.facebook.com/sharer/sharer.php?u=https%3A%2F%2F"e= "Share on Facebook") - [ ](https://twitter.com/intent/tweet?source=https%3A%2F%2F&text=:%20https%3A%2F%2F "Tweet") - [ ](http://www.linkedin.com/shareArticle?mini=true&url=https%3A%2F%2F&title=&summary=&source=https%3A%2F%2F "Share on LinkedIn") - [ ](mailto:?subject=&body=:%20https%3A%2F%2F "Send email") - **Community Partner:** [Korotu Technology]( https://www.korotu.com/ "Korotu Technology") - **Degree:** - Bachelor of Applied Science - **Program:** - [Computer Engineering](https://engineering.ubc.ca/programs/undergraduate/computer-engineering "Find out more about Computer Engineering") - **Campus:** Vancouver ## Our project We worked with [Korotu Technology](https://www.korotu.com/), a sustainable finance technology start-up, to develop ForestFolio, an application that enables non-technical users to quickly map the area below the forest canopy. The app captures data on tree location, diameter and height. > ForestFolio could enable larger areas of forest to be accurately measured in less time, providing decision-makers with the data they need to better manage forests. [Korotu Technology](https://www.korotu.com/) ## Our design solution Making evidence-informed decisions requires data. For forest management, that has traditionally involved sending trained surveyors into often remote forest locations to take measurements of tree height, diameter and location – a costly and time-consuming endeavour. More recently, drones and satellites have been used. Yet these technologies are unable to collect quality data below the forest canopy, consequently missing out on growth hidden underneath the tree cover. Our app, ForestFolio, simplifies and streamlines the data collection process. > It combines the tools used by traditional surveyors into one device, relying on a smartphone’s camera, LiDAR sensor, gyroscope and accelerometer to collect data. First, users walk around the forest plot and perform a LiDAR scan of their surroundings to generate a map of the detected trees within the plot area. Machine learning algorithms automatically detect trees with a diameter of more than nine centimetres, then estimate the diameter of each detected tree. Next, users are guided through a simple process for measuring tree height, which is calculated by measuring the angle from the ground to the top of the tree. This step includes an augmented reality feature to help users orient themselves and reduce error. This information is then uploaded to the server for additional analysis. ![COMP design poster](/sites/default/files/styles/max_325x325/public/2024-04/comp_poster.png.webp?itok=gB457gao) ## Our process > At the start of the project, we met with our client at a forested area here at UBC to learn how forest inventory data is traditionally collected. It took several hours for us to collect data for just 10 trees! Through our initial client meetings we developed a list of both essential and “nice to have” requirements for the app. > We conducted research to narrow down the best approach for each component of the system. For example, for the navigation component, we needed a solution that could function in a forest with no Internet connection that would enable users to track their location within a 400 m2 plot and know which trees still need to be scanned for data collection. We considered many options such as Tight Learned Inertial Odometry, which uses a neural network to estimate the change in device position over a specific time period. ## The challenges we faced There are some standalone apps that did elements of what we wanted, but very few were open source. This required us to do research, build our own knowledge base and develop our own methods. There were also challenges in finding the APIs and technologies that would enable us to achieve our goals, and iOS is notorious for not having support for many of the libraries we use. > Integrating each component into one application also had its challenges. In fact, we’re not sure if anyone has integrated these specific elements before! ## What we’re most proud of > It took months of research and debugging to get this working – and we had a lot of satisfaction when the different app components all came together and we were able to validate the results. We surprised ourselves with how well the tree diameter detection functioned. We’re also proud to have been able to meet one of the stretch goals, which was to measure tree height. The fact that we could incorporate that into ForestFolio is quite an achievement. ## Our project’s future > Our client was very happy with our work and ability to achieve the project goals. Features for future iterations would include the ability to work in a truly remote location (our version does require a GPS signal), collect data for different plot shapes and sizes, and automatically detect tree species. Additionally, the app could support a wider range of mobile devices, as the current version is limited to certain iPhone Pro devices that are equipped with the LiDAR sensor. [ UBC Electrical and Computer Engineering ![UBC Electrical and Computer Engineering logo](/sites/default/files/styles/max_325x325/public/2026-02/ubc-electrical-and-computer-engineering-logo-shortname-566x28.png.webp?itok=WQbzhJMi) ](https://ece.ubc.ca "UBC Electrical and Computer Engineering") - [Student Project](/spotlight-type/student-project "Student Project") - [Engineering](/topics/engineering "Engineering") - [International Student Experience](/topics/international-student-experience "International Student Experience") - [Student Experience](/topics/student-experience "Student Experience") - [Women in Engineering](/topics/women-in-engineering "Women in Engineering")